[Nagiosplug-help] Too many spawned nsca processes fill up proc table

[Ralph.Grothe at itdz-berlin.de]

Dear List Subscribers,

I run a Nagios 2.5 server on an AIX 4.3 box.

As I have quite a few passive checks from a distributed Nagios
server which checks hosts
within a firewalled LAN segment,
I set up an NSCA (2.4) daemon that is spawned by inetd.

$ grep nsca /etc/inetd.conf
nsca    stream  tcp     nowait  nagios  /opt/sw/nagios/bin/nsca
nsca -c /opt/sw/
nagios/etc/nsca.cfg --inetd
 
$ /opt/sw/nagios/bin/nsca -v 2>&1|head -5
Incorrect command line arguments supplied

NSCA - Nagios Service Check Acceptor
Copyright (c) 2000-2003 Ethan Galstad (nagios at nagios.org)
Version: 2.4


Though I am quite happy with this set up for most of the time
it nevertheless occassionally happens that inetd spawns so many
nsca processes
of incoming passive checks (or the nagios daemon isn't flushing
its external command fifo appropiately
so that the nsca processes are blocked on writing to the fifo?)
that the nagios user account gets its share of runnable user
space proc table so much filled up
that it cannot fork any processes anymore,
which renders the whole nagios server useless as it can't run any
checks anymore as well.
Of course in a situation as this the webinterface is unusable as
well for most of the tasks
(naturally if the external command fifo can't be read anymore)

Here's an example as has happened just now

$ ps -u nagios -o pid=|wc -l
     779
 
$ ps -u nagios -o args=|grep -c inetd
778
 
$ ps -u nagios -o args|head -2
COMMAND
nsca -c /opt/sw/nagios/etc/nsca.cfg --inetd 

# su - nagios
bash: fork: Resource temporarily unavailable


All I can do then is become root and kill all nagios processes,
and start nagios anew.

# kill $(ps -u nagios -o pid=)


Please, tell what is causing this (a possible bug?).
Has nsca been tested with AIX 4.3 inetd?

Would it alleviate the situation if I ran nsca stand-alone
instead spawned as inetd service?

Regards

Ralph